A comparison of six numerical methods for integrating a compartmental Hodgkin-Huxley type model

Abstract

We compare six numerical integrators' performance when simulating a regular spiking cortical neuron model whose 74-compartments are equipped with eleven membrane ion channels and Calcium dynamics. Four methods are explicit and two are implicit; three are finite difference PDE methods, two are Runge-Kutta methods, and one an exponential time differencing method. Three methods are first-, two commonly considered second-, and one commonly considered fourth-order. Derivations show, and simulation data confirms, that Hodgkin-Huxley type cable equations render multiple order explicit RK methods as first-order methods. Illustrations compare accuracy, stability, variations of action potential phase and waveform statistics. Explicit methods were found unsuited for our model given their inability to control spiking waveform consistency up to 10μs less than the step size for onset of instability. While the backward-time central space method performed satisfactorily as a first order method for step sizes up to 80μs, performance of the Hines-Crank-Nicolson method, our only true second order method, was unmatched for step sizes of 1−100μs.